Applied Urban Modelling 2012

CASA researchers were out in force at the AUM 2012 meeting in Cambridge last week, organised by the urban modelling group at the Martin Centre. It was an enjoyable meeting, with a good range of participants from both academia and built-environment practitioners. I’ll discuss some highlights from my own GIS and visualisation perspective.

It was great to see Paul Waddell present UrbanSim, which is a well established and popular open source platform for land use transport modelling based on microsimulation. Current improvements include adding 3D visualisation capabilities and pedestrian accessibility. Paul also had a demo of an impressive new urban design tool using a procedural architecture approach similar to CityEngine. Colleagues at CASA Camilo Vargas and Melanie Bosredon are developing an UrbanSim model of London, so we will be returning to this software in future posts.

Andres Sevtsuk from City Form Lab MIT presented on modelling retail locations from a street network Space Syntax type approach. His team have developed an Urban Network Analysis tool for performing measures like Betweeness and Closeness within ArcGIS. This tool is also open source and it’s great to see so much interesting software going down this free to access route.

My favourite presentation was from past and present CASA researchers Kiril Stanilov and Paolo Mascucci. Kiril has painstakingly been putting together an incredibly detailed vector dataset of the growth of London’s road network from the 1700’s to the present day. The time-lapse sequence of the network growing looked spectacular, highlighting the path dependence from historical forms and the different sequences of growth in London’s history. There’s fantastic potential in this dataset for improving modelling and understanding of how cities grow and develop. A flavour of the data can be seen in the below poster image:

Mapping Intra-metropolitan Journey-to-work Sustainability


As cities expand with multiple centres spread over massive regional hinterlands, the need to better understand the geographical variation across and within cities has become more pressing. This need applies strongly to issues of travel sustainability, where urban centres differ greatly in the accessibility they facilitate for private, public and active transport.

Spatial indicators are a useful tool to summarise complicated intrametropolitan patterns, as illustrated in my new working paper mapping CO2 emissions from journey-to-work travel across the London Region. The results of this indicator show a massive range of travel emissions by workplace of up to 300%, with particular problems for airports and the specialised employment region of the Western Sector, as can be seen in the map above.

This paper was co-authored with Joan Serras at CASA, who helped with the development of the road and public transport network analysis to model realistic routing behaviour from origin and destination flows from the 2001 census. One interesting aspect to this was the inclusion of GPS data to model average road speeds in London as illustrated below:

Full paper abstract:

“This paper develops a methodology for estimating network distances and CO2 emissions for UK census ward-level journey-to-work interactions. Improvements are made on existing empirical measures by providing comprehensive intra-metropolitan analysis; increasing network routing accuracy with UK public transport timetable and GPS-based average road speed data; allowing multimodal travel; and developing metrics suitable for travel sustainability analysis. The output unit of CO2 emissions has been selected to enable the integration of mode-choice and travel distance data, and to aid compatibility with integrated assessment applications. 

The methodology is applied to the case study of the London Region for the year 2001. A very high degree of intra-metropolitan variation is identified in the results. Employment sub-centres diverge in their per-capita CO2 emissions by up to 300%, with specific problems of carbon intensive commuting to major airports and the specialised employment region of the Western Sector. These findings indicate that subcentre travel variation may be intrinsic to polycentric urban structures. The paper discusses means to improve the methodology, in relation to issues of coefficient disaggregation and modelling more complicated multi-modal trips.”

Sponsored Space

Cities have always been the great spaces of commerce, trade and advertising; and recent decades have seen the corporate realm expand with the privatisation of services such as transport and utilities. Arguably corporate ownership is currently taking a new aspect with the explicit branding of urban places. English football clubs now play at the Emirates, Etihad and Sports Division Arena, rather than at Highbury, City of Manchester and St James’s Park stadiums. Skyscrapers in Canary Wharf project gigantic banking logos across East London. Westfield super-malls have branded entire new urban districts of streets and squares.

This process is not new- corporations have been building places to express their brands for a long time. Disneyland is perhaps the most famous example, and arguably the original UK new towns like Cadbury’s Bournville shared some of the same ideals. Many useful projects would simply not happen without corporate sponsorship, such as the popular (but loss-making)  Barclays Cycle Hire scheme in London, soon to be mimicked with Citi-Bike in New York.

The question however is whether this process represents a ‘slippery slope’ towards a comprehensive corporate branding of urban space. There has been much recent debate in London over the mayor’s decision to develop a cable car project across the Thames at the O2 Arena (another corporate venue). The project is sponsored by Emirates airlines and their name will be attached to two public transport stations, including appearing on London’s classic public transport map for a period of ten years.

The impact is fairly minor, though it does mean London is now in the dubious company of Dubai and Las Vegas in allowing sponsorship of subway stations. It did get me thinking, what would happen if London let rip and allowed the full sponsorship of the public transport network? My guess at the resulting network for London of businesses naming districts where they locate is below.

Obviously this map is satirical and would never actually be allowed to happen. I would argue though that this map does in its own way represent a real geography of corporations and capital flows in London. Is this map a more functionally honest representation of contemporary London than traditional images?